Re: status of http://www.w3.org/2002/03owlt/

Peter:
> This would be OK by me, as long as rejected tests were removed upon their
> rejection and as long as tests were only approved under the conditions
> above.

A peculiarity of W3C's web site is that content is not removed, ever.
This appears to be a non-negotaible site policy.

Given that the 404 error is (IMO) the ideal response when one tries to
access a test that has been rejected, I find this a peculiar policy.

Instead, if a test is rejected we are obliged to change the content to say
something like:

<!-- This was a rejected test case. -->

I share your desire for a clean deliverable from the test work.
I tend to agree with you that clearer and stricter process would help ensure
that, but I also agree with Dan that better process is not actually
necessary; if someone is prepared to do the work to tidy things up post hoc.

Possible machine generated clean test deliverables are:
+ a new directory with only approved tests in it
+ a document including inline all the approved tests
(this could be the bulk of an OWL Test Cases WD)
+ index files for all the directories in the test case directory that only
point to approved tests and hide rejected tests.

I think a process document might help, if I were to create one it would
essentially be a second draft of:
http://lists.w3.org/Archives/Public/www-webont-wg/2002Jul/0101.html

however I have no desire to try and police Dan and Jos - a thankless task!

Thus I tend to feel that the Test Cases WD with clear editorial
responsibility is the best way to end up with a clean test deliverable.

Jeremy

Received on Thursday, 29 August 2002 07:00:47 UTC